2021-11-07 A Qualitative Comparison of Zeiss Milvus 35mm 1.4 and Nikkor 35mm 1.8 Z, using the Nikon Z7
Disclaimer : I am no physicist, professional photography reviewer or lens designer specialist...this is NOT a scientific, complete nor a quantitative lens comparison study. It is possible there are errors in the methodology I used (you are welcome to tell me). I don’t know if the lens are perfectly aligned or not. 35mm Z is brand new. Milvus has been used for many years, but in perfect condition.
What I want to determine : I own Milvus 35mm 1.4 . Very happy, but amazingly heavy and huge. The question is : should I upgrade to a 35mm Z 1.8 S instead ? For me, most important criteria are, in order of importance :
Z7 Settings :
Shooting Methodology :
Processing Methodology :
What I want to determine : I own Milvus 35mm 1.4 . Very happy, but amazingly heavy and huge. The question is : should I upgrade to a 35mm Z 1.8 S instead ? For me, most important criteria are, in order of importance :
- no visible chromatic aberration in most circumstances (hate that…) at 100% zoom level at and above F1.8
- micro-contrast and « rendering », quality of « focus transition », subject isolation
- manual focus experience
- weight
- very good sharpness in general from F2, amazing from F4 from corner to corner
- Bokeh quality (not nervous, buttery)
- as little field curvature as possible (but don't know how to measure that...)
- color rendering
Z7 Settings :
- Active D_Lighting : OFF
- Vignetting Control : OFF
- Diffraction Compensation : OFF
- Auto Distortion Control : OFF
- Flicker Reduction Shooting : OFF
- Shooting in RAW only (lossless compression)
Shooting Methodology :
- Hand Held
- Pictures taken in the exact same lighting condition, exact same location when switching from 35mm S to Milvus 35, vice-versa, pictures taken just 1-2 minutes apart.
- I only used manual focus operated at maximum zoom level, great care has been taken to maximize accuracy of focusing. Same focusing point for both lenses.
- Professional gray card used for each lens shot
- Lenses are perfectly clean.
- Comparison has been done with exactly the same aperture, speed, ISO and camera settings
Processing Methodology :
- Raw Processor used is (the amazing…) Darktable 3.6.1 , Manjaro Linux. Settings are as minimalist as possible :
- Scene-referred workflow
- Filmic RGB ON : white=4.4 black=-7.7 contrast=1.35 latitude=33%
- Default demosaicing algo
- White balance has been adjusted for each picture, using a pro gray card.
- Exposure : to my surprise, 35mm S was consistently 0.5 EV less bright than Milvus. For the comparison, I added then 0.5 EV to all 35mm S images.
- All other modules are OFF (including sharpen, lens correction, chromatic aberration correction, distortion correction, vignetting correction)
Chromatic Aberration
Big difference here… From F1.8 to F2.8 (and even beyond), Milvus is clearly better, although not perfect neither for sure. No need to do pixel peeping to see it. Performance of the 35mm S is bad at these apertures. Not only it can easily be seen in high contrast regions, but the overall micro-contrast is sometimes lowered even in the center of the frame, casting an annoying purple halo around high contrast zones. Very disappointing. And yes, 100% zoom level is critical for me, as I always crop (sometimes significantly) my images. This is specially important for prime lens user.
Winner : Milvus (better to strongly better)
Big difference here… From F1.8 to F2.8 (and even beyond), Milvus is clearly better, although not perfect neither for sure. No need to do pixel peeping to see it. Performance of the 35mm S is bad at these apertures. Not only it can easily be seen in high contrast regions, but the overall micro-contrast is sometimes lowered even in the center of the frame, casting an annoying purple halo around high contrast zones. Very disappointing. And yes, 100% zoom level is critical for me, as I always crop (sometimes significantly) my images. This is specially important for prime lens user.
Winner : Milvus (better to strongly better)
Manual Focus Operation
Wow, what a surprise here...I was expecting fly-by-wire mechanism of the Nikkor to be problematic for precision manual focusing. Actually, I have been quite pleased with how efficient it is! The grip is extremely good (better than the rubber one of the Milvus). More surprising yet, I have found that I have more control with the Nikkor, because the « throw » is essentially infinite and selecting the rotation speed decides the amount of change for the focus. Superb !
Winner : Nikkor (better )
Wow, what a surprise here...I was expecting fly-by-wire mechanism of the Nikkor to be problematic for precision manual focusing. Actually, I have been quite pleased with how efficient it is! The grip is extremely good (better than the rubber one of the Milvus). More surprising yet, I have found that I have more control with the Nikkor, because the « throw » is essentially infinite and selecting the rotation speed decides the amount of change for the focus. Superb !
Winner : Nikkor (better )
Bokeh Quality
Again a subjective assessment. From what I can observe, Milvus is slightly better (« creamier », meaning less details in the background, less nervous). Visible even at F4. Probably linked in a way to the next item « subject isolation ».
Winner : Milvus (slightly better )
Again a subjective assessment. From what I can observe, Milvus is slightly better (« creamier », meaning less details in the background, less nervous). Visible even at F4. Probably linked in a way to the next item « subject isolation ».
Winner : Milvus (slightly better )
Subject isolation (3D pop...)
Ok, I don't like this term, it is too subjective and difficult to actually measure (what does it mean actually?). I never saw a study proving that a lens has better « suject separation » and more « 3D pop » than another one . This being said, for me, it should represents the ability to isolate the subject from the surrounding (background but also foreground). This is maybe equivalent to the so-called « dimensionnality » aspect. I have no idea if it makes any sense on the optical point of view that « focus transition » property is dependent on the lens optical formula (for a given focal length and aperture), please help me on that, but if it does, this could help separate the subject by having a rapid transition. Furthermore, it seems to me that a greater micro-contrast help to enhance the subject perception, hence its separation from a background that is more blurred. Anyway, I am just interested to the facts and what can be observed, whatever the cause behind. Look at this images below : focus is in the guy’s eyes :
Ok, I don't like this term, it is too subjective and difficult to actually measure (what does it mean actually?). I never saw a study proving that a lens has better « suject separation » and more « 3D pop » than another one . This being said, for me, it should represents the ability to isolate the subject from the surrounding (background but also foreground). This is maybe equivalent to the so-called « dimensionnality » aspect. I have no idea if it makes any sense on the optical point of view that « focus transition » property is dependent on the lens optical formula (for a given focal length and aperture), please help me on that, but if it does, this could help separate the subject by having a rapid transition. Furthermore, it seems to me that a greater micro-contrast help to enhance the subject perception, hence its separation from a background that is more blurred. Anyway, I am just interested to the facts and what can be observed, whatever the cause behind. Look at this images below : focus is in the guy’s eyes :
Clearly, it seems to me that the Milvus achieves a greater « subject separation ». (look at the building behind). Is it because the FOV of the Milvus is smaller ??? Cropped a little bit to show the same FOV, we have :
Milvus still better (look at the yellow sign upper right). And what about the sharpness and micro-contrast of the 2 people, the center of the focus ? Almost the same !
Here is another example at F4 this time,. Center is extremely sharp for both lenses, but once again we clearly see that the background is more « dissolved » by the Milvus : not by an enormous amount, but still a fact and having a visual impact
Hence I conclude, without being able to explain it, that Milvus offers a better subject isolation. Not by an enormous amount, but clearly visible and verifiable.
Winner : Milvus (better )
Winner : Milvus (better )
Sharpness
This is quite difficult to assess, since, assuming a flat field curvature, we need objects in the frame to be not too close (bye bye test charts) in order to reproduce field conditions, on the same plane and that plane must be parallel to the sensor… From F1.8 to F2.8, 35mm S is amazing, nothing less, it beats easily the Milvus, specially in the corners and edges where it is much better than the Zeiss. However, after F2.8, the test I have made indicates that Milvus becomes equal or superior to the Nikkor in the center (not in the extreme corners though). This is not small achievement.
Winner : Nikkor (slightly better overall, strongly better at large apertures)
This is quite difficult to assess, since, assuming a flat field curvature, we need objects in the frame to be not too close (bye bye test charts) in order to reproduce field conditions, on the same plane and that plane must be parallel to the sensor… From F1.8 to F2.8, 35mm S is amazing, nothing less, it beats easily the Milvus, specially in the corners and edges where it is much better than the Zeiss. However, after F2.8, the test I have made indicates that Milvus becomes equal or superior to the Nikkor in the center (not in the extreme corners though). This is not small achievement.
Winner : Nikkor (slightly better overall, strongly better at large apertures)
Vignetting
The Nikkor is not impressive as far as vignetting is concerned… From F1.8 to F4, the vignetting observed is high, worse than the Milvus, which is not spectacular either, but in the « average ».
Winner : Milvus (slightly better)
The Nikkor is not impressive as far as vignetting is concerned… From F1.8 to F4, the vignetting observed is high, worse than the Milvus, which is not spectacular either, but in the « average ».
Winner : Milvus (slightly better)
Overall conclusion
So… a tough call, neither lens is an absolute winner on all fronts.
The Nikkor is amazing on the sharpness point of view, at all apertures, including edges and corners, and easily beats Milvus from F1.8 to F2.8. It weighs much less than the Milvus, which is a very important factor to consider (aside from being much cheaper). Build and robustness is absolutely adequate. The manual focus is a real pleasure to use and unexpectedly superior in efficiency to Milvus (grip quality and variable degree of control).
The Milvus truely offers the famous « 3D pop » or whatever you call it, it is verifiable according to my tests. Not by a large amount though, but it makes a difference. From F2.8, sharpness increases dramatically and the Milvus begins to play in the same league than the 35mm S (except for corners). Amazing for a lens that old.
This being said, for me, the high amount of LoCA and also the very present vignetting from F1.8 to F4 on the 35mm S is too much for a high grade prime. It reminds me very much of the behavior of the 85mm 1.8 AF-S. LoCA is clearly visible in many usual shots and is not a mere artifact produced in rare circumstances. This in itself disqualifies the Nikkor 35mm 1.8S for a candidate to upgrade my Milvus 35mm 1.4
I will however follow very closely the expected arrival of the rumored 35mm 1.2 S. If this lens fixes the serious LoCA problems and lowers the vignetting, while keeping sharpness at a very high level corner to corner, with very low coma, then the Milvus will be replaced. One get spoiled with a Zeiss Milvus...cannot easily dial back the quality level when settling for a new lens.
So… a tough call, neither lens is an absolute winner on all fronts.
The Nikkor is amazing on the sharpness point of view, at all apertures, including edges and corners, and easily beats Milvus from F1.8 to F2.8. It weighs much less than the Milvus, which is a very important factor to consider (aside from being much cheaper). Build and robustness is absolutely adequate. The manual focus is a real pleasure to use and unexpectedly superior in efficiency to Milvus (grip quality and variable degree of control).
The Milvus truely offers the famous « 3D pop » or whatever you call it, it is verifiable according to my tests. Not by a large amount though, but it makes a difference. From F2.8, sharpness increases dramatically and the Milvus begins to play in the same league than the 35mm S (except for corners). Amazing for a lens that old.
This being said, for me, the high amount of LoCA and also the very present vignetting from F1.8 to F4 on the 35mm S is too much for a high grade prime. It reminds me very much of the behavior of the 85mm 1.8 AF-S. LoCA is clearly visible in many usual shots and is not a mere artifact produced in rare circumstances. This in itself disqualifies the Nikkor 35mm 1.8S for a candidate to upgrade my Milvus 35mm 1.4
I will however follow very closely the expected arrival of the rumored 35mm 1.2 S. If this lens fixes the serious LoCA problems and lowers the vignetting, while keeping sharpness at a very high level corner to corner, with very low coma, then the Milvus will be replaced. One get spoiled with a Zeiss Milvus...cannot easily dial back the quality level when settling for a new lens.
